The Campus AI Crisis 🔗 **Source:** [The Campus AI Crisis](http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html) > [!note]+ Document Notes > tags: [AI, higher-education, labor-market] > author: Jeffrey Selingo > source: New York Magazine (Intelligencer) > reference: [http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html] > date created: 2026-01-22 00:00 > > # 🏷️ Title: The Campus AI Crisis > <!-- deep-read:sections start --> ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) ### 🔗 References - Jeffrey Selingo, “The Campus AI Crisis,” New York Magazine / Intelligencer, 2026. - Stanford Digital Economy Lab study on Gen-Z and AI-exposed jobs. - Harvard/Elon/AAC&U studies on AI and early-career workers. - Burning Glass Institute, Strada, McKinsey Global Institute, NACE surveys. ## 📝 Summary - What happened: AI has rapidly reshaped entry-level jobs, especially in white-collar and tech-adjacent fields, blindsiding recent graduates and their colleges. - Why it matters: The traditional promise that a degree leads to stable, well-paid work is breaking down fastest for Gen-Z, eroding trust in higher ed. - Key stakeholders: Students and families, colleges and universities, employers, policymakers, and AI vendors. - Risks and tradeoffs: Over-indexing on “hot” majors; widening inequities by institution tier; slow academic governance vs. fast labor-market shifts. - Bottom line: Colleges must integrate work-based learning and AI fluency into the core of undergraduate education or risk irrelevance in an AI-dominated labor market. ## 📖 Main Content - Core arguments: - AI is hitting *junior* knowledge workers hardest, reversing the usual pattern where young employees benefit from new tech. - Entry-level roles in fields like software, data, and customer support are being automated or re-scoped, shrinking opportunities for new grads. - Higher ed still operates as if classroom instruction is central and work is peripheral, even as internships and applied experience are now decisive. - AI literacy alone is not enough; enduring advantages are judgment, critical thinking, communication, and the ability to work *with* AI. - Institutions must structurally embed work-based learning and AI fluency (e.g., Ohio State’s AI Fluency initiative) rather than treating them as add-ons. - Context & terms: - “Tacit knowledge”: undocumented, experience-based know-how that helps senior workers outperform AI and juniors. - “Underemployment”: working in jobs that do not require a college degree. - “Work-based learning”: internships, co-ops, apprenticeships, company projects tied to coursework. - Uncertainties / contested: - Exact share of early-career job losses attributable to AI vs. macroeconomic conditions. - Whether AI will ultimately create more/better jobs than it destroys, and on what time horizon. - How much colleges can realistically retool, given governance, financial, and political constraints. ### 🗒️ Notes - Author frames article with his own experience entering the workforce during the early internet to parallel today’s AI transition. - Case: Alina McMahon, Pitt marketing major with strong internships (including Paramount), faces ~150 unanswered applications; employers cite AI/outsourcing for cuts ([McMahon’s job search](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=00008_00000-008R&startLexposition=00006&emphasizeText=applied+to+roughly+150+jobs)). - Data: 15% fewer entry-level/internship postings (Handshake, 2025) and 26% more applications per posting; new-grad unemployment 5.7%, above national average and even HS grads. - Handshake’s Cruzvergara: employers are *choosing* to automate pieces of jobs instead of redesigning roles around people. - Stanford Digital Economy Lab: Gen-Z college grad employment in AI-affected jobs down 16% since late 2022; senior workers buffered by tacit knowledge and better ability to spot AI hallucinations. - Harvard study: less experienced workers more affected; mid-tier college grads see steepest employment drop—too costly vs. skills compared to elite and lower-tier peers ([mid-tier squeeze](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000E_00004-00A4&startLexposition=0000E&emphasizeText=Mid-tier+graduates+end+up+somewhat+in+between)). - Fed chair Powell flags concern for new grads; NACE survey shows unusually pessimistic employer outlook for class of 2026. - Colleges reacted first by treating generative AI primarily as a cheating threat: bans, blue books, blocked tools ([AI as cheating tool](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000I_00002-001Z&startLexposition=0000G&emphasizeText=world%E2%80%99s+greatest+cheating+tool)). - Now realize AI is creating a structural gap between what degrees are *for* and what the labor market demands; students absorb the risk. - Structural shifts: post-2008, students pivoted away from humanities toward “employable” majors like computer science under a “get a job” mindset. - Comp-sci enrollment more than doubled (to 170k majors) even as total undergrad numbers fell—driven by Code.org, Girls Who Code, political leaders, and tech luminaries endorsing coding. - Irony: young CS grads now among those with highest unemployment; entry-level coding roles are among the most AI-exposed; CS enrollment now dropping ~8% year-on-year ([CS whiplash](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000Q_00008-0010&startLexposition=0000Q&emphasizeText=College+graduates+in+their+20s+with+computer-science+and+computer-engineering+majors+have+one+of+the+highest+unemployment+rates)). - Case: Andrew Wyatt leaves CS for econ/data science after seeing AI do most of his prior internship tasks; now even data-centric roles are “AI can do it all” ([AI can do it all](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000S_00000-00Do&startLexposition=0000S&emphasizeText=AI+can+do+it+all+without+me)). - Work experience as durable differentiator: internships strongly predictive of avoiding underemployment; majority of non-interns underemployed 5 years later ([internship effect](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000U_00004-002v&startLexposition=0000U&emphasizeText=More+than+half+of+graduates+who+did+not+intern+as+students%2C+according+to+Burning+Glass%2C+were+underemployed+five+years+after+college)). - Only a minority of institutions structurally integrate co-ops/internships (Northeastern, Drexel, Cincinnati); most still valorize classroom learning and marginalize career centers, which many students never use. - Ryan Craig: colleges face an existential issue—must make relevant, in-field, preferably paid work central for *every* student, often via intermediaries that broker placements. - Workday CLO Chris Ernst: 70% learning from experience, 20% from relationships, 10% from formal instruction ([70-20-10 rule](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000g_00004-000o&startLexposition=0000e&emphasizeText=70+percent+of+learning+comes+from+experience)). - McKinsey: demand for AI skills in job postings up ~7x in two years, fastest-growing skill signal. - AAC&U/Elon survey: <50% of college leaders say campuses are ready to use AI for student preparation; many students report AI was ignored or discouraged in classes and internships. - Faculty dinners: movement from blanket bans to course-level norms; early adoption of AI for “messy work” (summaries, study guides, drafts) potentially freeing time for higher-order skills. - But curricular change is slow; faculty incentives to experiment are weak; teaching practices are path-dependent (“one-trick ponies”). - Elite employers (consulting, Wall Street, law) demand both AI literacy and strong human judgment; Carlyle example: AI speeds research, but humans must write assessment paragraphs to demonstrate understanding. - Higher ed faces intersecting crises: finances, enrollment, politics, research funding, international enrollments—limiting capacity to focus solely on AI and work. - Ohio State’s AI Fluency initiative aims for 2029 grads to be fluent in AI *in their field*, beyond mere literacy; baseline across university plus major-specific depth ([AI Fluency 2029](https://read.readwise.io/read/01kfk3mvybrg6f2ebbr4zjz3q3?endLexposition=0000u_00002-00B9&startLexposition=0000u&emphasizeText=class+%E2%80%94+this+year%E2%80%99s+freshmen+%E2%80%94+that+the+university+has+promised+will+be+%E2%80%9Cfluent)). - Business school example: Prof. Vince Castillo’s “Amsterdam for AI” class—open AI use with transparency; non-coding business students leverage AI to interrogate data and think more strategically. - Historical analogy: colleges first treated the internet as encyclopedia/email, then gradually rewired teaching and programs; AI is likely to follow a similar “tool to infrastructure” trajectory. - Lindsey Pollak: we’re bad at predicting which majors/skills matter; steering students toward supposedly “safe” skills (like coding) can backfire when tech shifts. - Closing note to parents: future is ambiguous; learning to navigate ambiguity itself is perhaps the key skill in an AI-accelerated world. ## 📚 References - “The Campus AI Crisis,” New York Magazine / Intelligencer, Jeffrey Selingo (primary article). - Stanford Digital Economy Lab – studies on AI-exposed occupations and Gen-Z employment. - Harvard University / Seyed Mahdi Hosseini Maasoum – AI and employment by institution tier. - Burning Glass Institute – underemployment and internships. - Handshake, NACE – early-career job market surveys. - McKinsey Global Institute – AI skill demand trends. - Ohio State University – AI Fluency initiative announcements. ## 🗣️ Paraphrase - Paragraph A: AI is shrinking and reshaping entry-level white-collar jobs, leaving recent graduates—especially in once-“safe” majors like computer science—without the roles they were trained for, while colleges lag in updating programs. - Paragraph B: To stay relevant, universities must embed AI fluency and real-world work experience into the heart of undergraduate education and help students develop enduring skills—judgment, critical thinking, communication—that complement AI in an uncertain labor market. ## 📰 Newsletter In this week’s WordPress-adjacent higher-ed briefing, we zoom out from plugins and performance to the pipeline itself: the grads who’ll build tomorrow’s sites and AI-powered experiences. Jeffrey Selingo’s “The Campus AI Crisis” tracks how generative AI is erasing many traditional entry-level tech roles while colleges cling to lecture-centric models. For the WordPress ecosystem, this raises two questions: how do we onboard new contributors when AI handles so much boilerplate, and what skills should WP-focused programs teach? The article suggests a mix: real projects, apprenticeships, and AI-augmented judgment over rote coding. Think: student teams shipping micro-plugins, performance audits, or content workflows using tools like Playground, Tide, and AI-assisted editors—guided by maintainers who emphasize architecture and ethics, not just syntax. If you run a dev shop, university program, or meetup, consider building structured internships around WordPress: data-informed theme optimization, accessibility sprints, or AI content QA for campus sites. Done right, we can make the WordPress community a proving ground where new developers learn to work *with* AI—and find the kind of work experience the current degree pipeline often fails to provide. ## ✅ Action Items #### DevRel - Map AI + WordPress skills (prompting, code review, content workflows) into sample curricula and workshop outlines for university partners. - Build AI-ready quickstarts: Playground-based labs where students ship a simple plugin/theme using AI assistance and human review. - Create a talk/demo series for campuses: “From AI Boilerplate to WordPress Judgment Calls” with concrete examples and repos. - Add case studies showing how interns can add value on WordPress performance, accessibility, and content operations with AI in the loop. - Instrument tutorials and SDK examples with metrics to track student progression from copy-paste to critical evaluation of AI output. - Launch a student-focused DevRel channel (Discord/Matrix) with office hours around WordPress + AI careers and contribution paths. #### OSPO - Define contribution policies for AI-assisted code in WordPress-related repos (disclosure, review standards, provenance). - Establish license-compliance and content-provenance rules for AI-generated artifacts (code, docs, media). - Integrate AI-related security checks (prompt injection, poison patterns) into vulnerability intake and SBOM workflows. - Maintain an inventory of AI-reliant components in the stack (CI helpers, doc tools) with ownership and review cadence. - Run training for maintainers on evaluating AI-generated patches and documentation changes. #### Community - Host “first contribution with AI” pairing sessions, emphasizing review, testing, and documentation—not just code generation. - Offer roadmap clinics where maintainers identify tasks suitable for students and AI-augmented contributors. - Publish transparent guidelines for sponsorships and stipends tied to teaching, mentoring, and apprenticeship-style work. - Practice cross-project incident drills involving AI-generated regressions or documentation failures. #### WordPress - Core: Label “AI-friendly” issues suitable for juniors using AI, but require tests and rationale in their own words. - Core Performance: Develop dashboards and experiments where students use AI to propose optimizations that are then benchmarked. - Core AI: Define responsible-usage patterns, evaluation sets, and UX guidelines for AI features in the editor and beyond. - **Core Program Team:** Coordinate with education partners so student projects line up with real roadmap needs and support bandwidth. - Mobile: Create AI-assisted bug triage flows while preserving clear UX review gates. - Playground: Build browser-based labs that simulate real-world client scenarios (performance, SEO, accessibility) with AI suggestions. - Community: Provide meetup kits on “Teaching with WordPress and AI” for educators and organizers. - Training: Publish learning paths that culminate in shipping a real fix or feature to .org using AI as a helper, not a crutch. - Marketing: Tell stories of students and early-career devs succeeding via WordPress apprenticeships and co-ops. - Support: Document best practices for AI-assisted support responses and escalation. - Design: Offer examples of AI-generated design explorations reviewed by human designers to teach critique skills. - Documentation: Add examples-first patterns demonstrating how to verify AI-suggested code against official docs. - Photos/TV/Openverse: Clarify policies around AI-generated media, captions, and licensing in the WP ecosystem. - Polyglots: Experiment with AI-assisted translation drafts followed by human review, documenting effective workflows. - Accessibility: Produce guidance on AI tools that help find issues vs. those that risk inaccessible output; keep humans in the loop. - Meta/CLI/Hosting/Tide/Themes/Plugins: Define AI usage and review guidelines across infrastructure, scoring, and reviews to protect ecosystem quality. #### The WPOCC - Pilot microgrants and fellowships for student teams building WordPress-based educational tools and AI literacy resources. - Fund documentation and “how to mentor AI-era contributors” guides for maintainers. - Support cross-institution collaborations where multiple campuses share WordPress-based curriculum pieces and labs. - Experiment with milestone-based funding for apprenticeships and co-op programs centered on WordPress projects. ### FAIR - Encourage local AI tooling for contributors, with clear provenance for generated code and content. - Publish governance templates that define AI usage norms, funding transparency, and anti-capture clauses for sponsored AI tooling. - Turn working agreements (code of conduct, AI guidelines, contribution policies) into reusable governance “products” with scorecards. ## 🧪 Hypotheses - H1: Embedding structured, paid work-based learning into undergraduate programs significantly reduces underemployment rates for graduates in AI-exposed fields. - H2: Students trained in AI fluency plus critical thinking and communication outperform peers with primarily technical training on early-career outcomes. - H3: Mid-tier institutions that rapidly adopt AI fluency and co-op models will narrow the employment gap with elite institutions. - H4: AI-assisted coursework without explicit reflection (e.g., summary paragraphs) leads to weaker long-term skill retention than AI + structured reflection. - H5: Intermediaries that broker apprenticeships at scale increase employer willingness to hire entry-level talent despite AI. ### 📊 Methodology - Longitudinal cohort studies comparing graduates from co-op-heavy vs. traditional programs on employment, salary, and underemployment. - Randomized controlled trials in courses: AI-ban vs. AI-allowed-with-reflection vs. traditional, measuring learning, internship offers, and job outcomes. - Comparative institutional analysis: mid-tier colleges adopting AI fluency initiatives vs. peers, tracking graduate employment over 5–10 years. - Employer surveys and hiring data analysis on AI skill keywords, internship experience, and retention. - Learning analytics combining LMS data, AI tool logs, and performance assessments to model skill gains. ### ❓ Questions - How can universities change governance quickly enough to keep curricula aligned with fast-moving AI labor markets? - What mix of generalist vs. specialist AI skills best protects early-career workers from automation? - How should families and advisors counsel students on majors given high uncertainty about future job categories? - What AI guardrails are needed so student use builds judgment instead of eroding it? - Can open-source ecosystems like WordPress function as de facto apprenticeship platforms at scale? ### 🔮 Future Research - Long-term impacts of AI on career ladders when traditional entry-level roles shrink or vanish. - Models of higher ed where “work-first, degree-later” (apprentice nation) is the default in digital fields. - Effective ways to teach tacit knowledge and error-detection in AI outputs. - Institutional strategies that help mid-tier colleges avoid being squeezed between elite and low-cost competitors. - Cross-national comparisons of AI-driven higher-ed reforms (e.g., UK/Australia intermediaries) vs. US approaches. ### 💡 Implications - Practice: Educators should integrate live client projects, AI-augmented workflows, and reflection assignments into courses. - Policy: Governments may need to incentivize apprenticeships and co-ops and rethink funding tied solely to seat time and degrees. - Security: AI-generated code and content in production require new review, provenance, and monitoring practices. - Education: Advising must foreground ambiguity tolerance, career experimentation, and skills that transfer across roles. - Funding: Philanthropy and public funds could prioritize intermediaries and infrastructure that de-risk early-career hiring. ### 🧠 Learning Styles Assessment - Reading level: Approx. 11th–12th grade (Flesch–Kincaid) given sentence complexity and economic/policy vocabulary. #### 👀 Visual learners - Present: Narrative descriptions, stats, and case examples (graphs not included in text). - Improve: Add charts on unemployment trends, flows between majors, and timelines of AI/tool adoption; course diagrams showing AI in workflows. #### 🔊 Aural learners - Present: No built-in audio; argument is text-only. - Improve: 3–5 minute narrated summary and panel discussions with students, employers, and faculty; podcast segment on AI and entry-level work. #### 🗣️ Verbal learners - Present: Rich narrative, quotes, and clear explanations; contrasting viewpoints from multiple experts. - Improve: Sidebars defining key terms (tacit knowledge, underemployment, AI fluency) and analogy-driven explanations. #### 🧩 Kinesthetic learners - Present: Conceptual descriptions of co-ops and internships, but no hands-on activities. - Improve: Interactive labs where students simulate job searches, design AI-augmented workflows, or draft AI policies. #### 🧠 Logical learners - Present: Data points, causal reasoning about AI and labor, and structured argument progression. - Improve: Tables comparing institutional models; decision trees for advising students about majors under uncertainty. #### 🧑‍🤝‍🧑 Social learners - Present: Stories of individuals and discussions among leaders; implied need for dialogue. - Improve: Guided seminar questions, group projects designing AI-ready curricula, peer review of AI use norms. #### 🧘‍♀️ Solitary learners - Present: Article suitable for solo reflection and note-taking. - Improve: Reflection prompts about career planning in an AI world; self-assessment checklists for AI readiness. #### 🌿 Naturalistic learners - Present: Minimal reference to environmental/sustainability topics. - Improve: Analogies to ecosystems (labor markets as shifting habitats), and links to energy and environmental impacts of AI infrastructure. ### 📊 Accessibility Assessment #### Screen reader semantics - Present: If rendered as typical web article, headings and paragraphs likely available. - Improve: Ensure semantic HTML5 landmarks, descriptive alt text for any figures, and accurate language attributes. #### Keyboard navigation - Present: Standard article likely keyboard-accessible. - Improve: Confirm visible focus states, logical tab order, skip links for nav. #### Contrast and readability - Present: Depends on publisher’s design but typically adequate. - Improve: Verify 4.5:1 contrast, comfortable line heights, and no text embedded in low-contrast images. #### Text resizing and reflow - Present: Modern responsive layout usually resizes. - Improve: Test at 200% zoom and high zoom on mobile; ensure no clipping or horizontal scroll for body text. #### Media and captions - Present: Primarily text; minimal media. - Improve: Add audio version with transcripts and any videos with captions and audio descriptions. #### Forms and interactions - Present: Likely minimal (newsletter signups, etc.). - Improve: Ensure labels, accessible error messages, and adequate target sizes. #### Standards reference - Map improvements against WCAG 2.2 AA, particularly 1.4 (distinguishable), 2.1 (keyboard), 2.4 (navigable), 3.3 (input assistance). ### 🌍 DEIB Assessment #### Cultural breadth - Present: Focus on US higher ed and labor markets; examples from mainstream institutions and employers. - Improve: Include international perspectives on AI in higher ed and labor, plus non-US case studies. #### Inclusive language - Present: Generally neutral, avoids stereotypes; references “Gen-Z,” “mid-tier” institutions with care. - Improve: Clarify that institutional tiering is descriptive, not value-laden; provide more definitions to reduce jargon barriers. #### Neurodiversity support - Present: Clear structure but long-form prose may be demanding. - Improve: Break content into shorter sections, add summaries and bullet lists; provide predictable navigation and callouts. #### Representation - Present: Mix of student, employer, and institutional voices; not heavily focused on demographic diversity. - Improve: More quotes from students of varied backgrounds, community colleges, and non-traditional learners. #### Psychological safety - Present: Acknowledges anxiety and discouragement among students in a matter-of-fact way. - Improve: Provide explicit resources or suggestions for coping, advising, and alternative pathways to reduce doom narratives. ### ✅ Logical Validity #### Strong reasoning - Inductive: Uses multiple data sources (Handshake, Fed, Stanford, Harvard, McKinsey) to infer patterns about AI’s impact on early careers; acknowledges uncertainty. - Causal: Carefully links AI adoption to job redesign/automation while noting that macroeconomics and structural changes also contribute. - Analogical: Compares AI’s trajectory to the internet’s adoption in higher ed, illustrating likely long-run transformation. #### Fallacies - Potential survivorship/selection risk: Focus on certain majors and institutions may not generalize to all students; could be made explicit. - Possible hindsight bias in critiquing push toward coding; partially countered by quoting experts about unpredictability. Fix: Add more explicit caveats about sample limitations and alternative explanations for observed trends. #### Tone of voice - Tone: Concerned, analytical, and empathetic toward students; mildly critical of institutional inertia but not hostile. - Uses personal anecdotes plus data to balance narrative and evidence. ### 🛡️ DARVO Analysis - Deny: The article does not deny student struggles; it foregrounds them. - Attack: It does not attack students; it offers mild critique of colleges and employers but not in a scapegoating way. - Reverse victim/offender: No attempt to portray powerful actors as victims; instead, students are shown as bearing disproportionate risk. - Overall, DARVO dynamics are not present; the piece is more diagnostic than defensive. ## 🎯 Conclusion AI is rapidly reshaping entry-level jobs, particularly for college-educated Gen-Z, while higher education’s slow-moving structures leave many unprepared. The article argues that colleges must re-center around work-based learning and AI fluency, while doubling down on timeless human skills like judgment and critical thinking. For students, families, and ecosystems like WordPress, the practical next step is to treat ambiguity as a given and design pathways—internships, apprenticeships, real projects—that build experience plus AI-augmented discernment, not just credentials. ## 🧱 Refactor Appendix (append after 🎯 Conclusion) ### 📚 Literature Split Pad - Purpose: Stage Konik CEE units that become per-source Literature notes. - Split level: **H3**. - Include verbatim quotes under 25 words and a [view highlight] link. ### [H1] AI disruption hits entry-level graduates first > [!quote] [The Campus AI Crisis](http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html) by [[Jeffrey Selingo]] via [[New York Magazine (Intelligencer)]]. Published 2026. Accessed 2026-01-22. > “Employment for Gen-Z college graduates in AI-affected jobs…has fallen by 16 percent since late 2022.” Two lines: Early-career roles in AI-exposed occupations are shrinking fastest, leaving new grads with fewer on-ramps. Use for arguments about rethinking entry-level design. [view highlight](http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html#H1) #xref #articleseed ID:rw_[MISSING] ### [H2] Work-based learning is now essential, not optional > [!quote] [The Campus AI Crisis](http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html) > “They need to figure out how to integrate relevant, in-field, and hopefully paid work experience for every student.” Two lines: Co-ops, internships, and apprenticeships have become central to employability in an AI labor market. Use when motivating structural reform in higher ed. [view highlight](http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html#H2) #xref ID:rw_[MISSING] --- ### 🔖 Atomic Split Pad (Zettel Candidates) - Purpose: Promote reusable claims into Atomic notes. - Split level: **H4**. - Each item: one idea, present tense, specific, testable. #### [Z1] AI disproportionately harms early-career knowledge workers - Scope: Gen-Z college grads in AI-exposed jobs since 2022. - Pointer: Stanford Digital Economy Lab findings summarized in Selingo (2026). - Links: [[AI-and-labor-markets]] [[Early-career-transitions]] #### [Z2] Universal work-based learning reduces graduate underemployment - Scope: Institutions embedding internships/co-ops for all students. - Pointer: Burning Glass Institute and Strada data via Selingo (2026). - Links: [[Work-based-learning]] [[Higher-ed-reform]] --- ### ✏️ Fleeting Queue (Extract by Selection) - Purpose: Capture quick actions or sparks for Fleeting notes. - Extraction: **Select one line** → “Refactor: Extract selection to new note (content only).” - Keep each line short and self-contained. - [F1] Draft a campus workshop outline on AI fluency plus internships—http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html #fleeting #inbox - [F2] Map WordPress contributor roles to “work-based learning” for CS students—http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html #fleeting #inbox - [F3] Design a rubric for evaluating AI-assisted student work focusing on judgment—http://nymag.com/intelligencer/article/what-is-college-for-in-the-age-of-ai.html #fleeting #inbox --- <!-- deep-read:sections end -->